OPTIMIZATION OF TiN COATED CARBIDE REAMING PROCESS ON SS 440C USING GREY
RELATIONAL ANALYSIS
Abstract— Manufacturing techniques in engineering fields, especially in the aerospace applications and the defence field are advancing day by day. While considering the manufacturing of any components of aerospace industry it is very important to increase the production quality associated with its manufacturing. The manufacturing of Electro Hydraulic Servo Valve (EHSV) considered for this work is a two stage electrically operated hydraulic valve, in which the output flow is proportional to the input current. Here investigated on both theoretical and practical aspects of using different processes involved in the production of EHSV. The material used for EHSV is SS 440 C. This investigation gave me a complete idea about the problems related to different machining processes associated with the manufacturing of EHSV, especially the problems related to the wire electric discharge machining (WEDM). The problems associated with WEDM including its high surface roughness value and high cylinricity for manufacturing cylindrical bores can only be eliminated by replacing the existing WEDM process with suitable process which produce components with better quality. For this purpose TiN coated carbide reamer is the best option to overcome the above mentioned problems for the manufacturing of cylindrical bores. In this work the analysis of existing process, optimization of proposed process, statistical and experimental comparison between the existing and proposed processes including response comparison are included. This study investigates the effects of various parameters such as speed, feed and allowance on the surface finish and cylindricity of the EHSV valve body. Grey Relational Analysis is used to optimize the proposed reaming process. Confirmation tests were performed to validate the results. Digital surface profilometer images were taken in order to investigate the surface characteristics.
Keywords— ANOVA, Cylindrical bores, EHSV, Grey relational analysis, Mean MRPI plot, Taguchi’s design of experiments, TiN coated carbide reamer
I. INTRODUCTION
The Electro Hydraulic Servo Valves (EHSV) is basically an electrically operated valve in which the output flow is proportional to input current. The EHSV is a two-stage flow control valve, in which the output flow is proportional to the input current. Different varieties of EHSV are available according to the requirements. The main difference between different EHSV is in its flow rates. The EHSV works in conjunction with an Electro Hydraulic Servo Actuator which is essentially a double acting type actuator. The actuator moves the Fin Tip [5] (controlled elements) with reference to the given command and operates in a closed loop. The material used for manufacturing EHSV is SS 440 C.
Figure 1. Cross section of EHSV
Figure 1 shows the cross section of an EHSV. There are so many manufacturing techniques used in precision manufacturing of components. Wire EDM is most extensively use techniques in manufacturing components with intricate shapes and profile.
But this process is not suitable for manufacturing components with cylindrical bores and uniformly tapered bores, as they are time consuming and costly process. Application of advanced reamers like TiN coated carbide reamer can be used to overcome the problems related to manufacturing of cylindrical and tapered bore manufacturing of components in precision manufacturing. Technological and economical comparison, surface integrity, corner error, crater, corrosion, etc.
of WEDM process where studied with the help of literatures [1], [2], [3], [4], [6], [7], [8], [9], [10].
While considering the manufacturing of EHSV, the WEDM is applied for the manufacturing of cylindrical bores in the EHSV valve body. But it is a time consuming and costly process. Also WEDM has several disadvantages as seen from several journals. So advanced process such as special purpose reaming process can be used instead of WEDM process for cylindrical bore manufacturing. Reaming is common machining process for enlarging, smoothing and accurately sizing existing holes to tight tolerances [5]. International manual [14] for special purpose tools gives a complete idea about the type of tools that suitable for different work materials. In that manual the type of reamer suitable for machining SS 440C is found as TiN coated carbide reamer. It was found that the most influencing parameter for reaming process are speed, feed, depth of cut and allowance. In this work, depth of cut is not considered as the variable parameter; as it can only be varied according to the length of spool bore. Reaming was retained for finishing, as a process capable of yielding required results when performed on inexpensive machine tools such as drill press with simple fixtures. Reaming is suitable for batch production typically of job shops, facing the challenges of meeting specifications at competitive costs.
Better surface finish and bore geometry can be obtained with the help of reaming process. Apparently minor influences led to enhanced process control and substantially better results at no extra cost. Results supported selection of production parameters meeting specified quality and cost targets, as well as substantial improvements [11]. From [13] it was found that average surface roughness for reaming range between 0.8 µm and 3.2 µm, but high-accuracy reaming can produce average surface roughness as low as 0.4 µm. The quality of the reamed holes was evaluated in terms of geometrical characteristics (diameter, roundness, cylindricity and surface roughness). Roundness as well as cylindricity was verified to be smaller than the specified tolerance. Higher feed rate leads to lower and more repeatable roughness. MQL in reaming leads to high quality results in terms of hole dimensions and surface finish. [12] The feed and speed are important process parameters to control surface roughness, tool wear, material removal rate and hole diameter error.
Thus it is essential to employ suitable combination of cutting speed and feed so as to reduce the variations that can affect the quality of the holes.
II. EXPERIMENTSANDMETHODS
The main objective of this project work is to find out how various reaming parameters such as speed, feed and allowance influence various output characteristics such as surface finish and cylindricity while considering the manufacturing of cylindrical bores in the EHSV valve body.
Figure 2. EHSV valve body
The valve body is the main component of EHSV. Figure 2 shows the EHSV valve body. When consider the valve body, the spool bore is the most important part of valve body. The spool bore is the only port where spool movement takes place. So while considering the valve body manufacturing, the manufacturing of spool bore was considered specifically and the manufacturing of other ports of valve body generally considered.
Figure 3. Bores in EHSV valve body
A. Existing Wire EDM Process
Wire Electrical Discharge Machining (WEDM) is a widely accepted non-traditional material removal process used to manufacture components with intricate shapes and profile. It is a specialized thermal machining process capable of accurately machining parts with varying hardness or complex shapes that are very difficult to machine by the main stream machining process. The basic features of WEDM are shown in figure 4. The principle of the WEDM process is based on the conventional EDM sparking phenomenon utilizing the widely accepted non – contact technique of material removal. However, WEDM utilizes a continuously travelling thin wire electrode made of copper, brass, or tungsten of diameter 0.03-0.30 mm, which is capable of achieving very small corner radii. The wire and there is no direct contact between the work piece and the wire, eliminating the mechanical stresses during machining. In addition, WEDM process is able to machine exotic and high strength and temperature resistive (HSTR) materials and eliminate the geometrical changes occurring in the machining of heat-treated steels.
Since the introduction of the process, WEDM has evolved from a simple means of making tools and dies to the best alternative of producing micro-scale parts with the highest degree of dimensional accuracy. WEDM has greatly altered the tooling and manufacturing industry, resulting in dramatic improvements in accuracy, quality, productivity and profit.
Over the years, the WEDM process has remained as a competitive machining option fulfilling the demanding machining requirements imposed by the short product development cycles and the growing cost pressures. However, the risk of wire breakage and bending has undermined the full potential of the process drastically reducing the efficiency and accuracy of the WEDM operation. A significant amount of research has explored the different methodologies of achieving the ultimate WEDM goals of optimizing the numerous process parameters analytically with the total elimination of the wire breakages and improving the overall machining reliability which are explained in the literature survey.
Figure 4. Basic features of WEDM
B. Experimental details for proposed reaming process
The experiment was conducted on vertical milling machine as shown in figure 5. The machine has speed settings up to 15,000 RPM, feed settings up to 30,000 mm/min. The experiments are conducted on stainless steel 440C.
Figure 5. Experimental reaming setup
Figure 6. TiN coated carbide reamer
The number of experiments and input levels are decided based on the design of experiments and the input parameters and their levels are presented in table 1.
The responses that needed to be taken into consideration are surface roughness and cylindricity which taken with Taylor Hobson Talysurf & Taylor Hobson Talyrond. In order to achieve the objectives, the experiments were designed based on Taguchi’s design of experiments.
Table 1: Reaming Prameters
Input Parameters Speed (RPM) Feed (mm/min) Allowance (mm)
Level 1 600 80 0.2
Level 2 800 120 0.3
Level 3 1000 160 0.4
Design of Experiments
Experimentation can be used for developing new product/processes as well as for improving the quality of existing product/processes. DOE is a systematic approach to investigation of a system or process. A set of input variables (factors) of a process or system are tested according to the design. In any experimentation the investigator tries to find out the effect of input variables on the output/performance of the product/process. To test the effect of these factors on a response variable, a suitable experiment is designed such that the necessary data for testing the significance of the effects of the factors on the response variable are collected. Identification of the response variable is the first step to be followed for the designing of experiment. Identify the factor affecting response variables, fix the number of levels of each factor, form the skeleton of the experiment and define its components comprises the next step for designing an experiment.
Overview of Taguchi Method
Taguchi method involves reducing the variation in a process through robust design of experiments. We can produce high quality products with minimum cost to the manufacturer by using Taguchi method. Taguchi developed a method for designing experiments to investigate how different parameters affect the mean and variance of a process performance characteristic that defines how well the process is functioning. The experimental design proposed by Taguchi involves using orthogonal arrays to organize the parameters affecting the process and the levels at which they should be varies. L9 orthogonal array is selected with the help of MINITAB 16 software.
Table 2: L9 Orthogonal Array
Experiment P1 P2 P3
1 1 1 1
2 1 2 2
3 1 3 3
4 2 1 2
5 2 2 3
6 2 3 1
7 3 1 3
8 3 2 1
9 3 3 2
These data will help to determine the factor which affects the product quality with minimum number of experiments.
Instead of having to test all possible combinations like the factorial design, the Taguchi method tests pairs of combinations. This allows for the collection of the necessary data to determine which factors most affect product quality with a minimum amount of experimentation, thus saving time and resources. The Taguchi method is best used when there are an intermediate number of variables (3 to 50), few interactions between variables, and when only a few
variables contribute significantly. Following is the array lookup table developed by Taguchi based on the number of parameters and levels.
Scheme of Experiments
Based on L9 orthogonal array experiments are conducted on Stainless steel 440C with TiN coated carbide reamer tool for different experiment levels which are shown in Table 1. Speed, feed and allowance are selected as input variable parameters. Three varying parameters with three different levels are taken for the purpose of design of experiments.
Table 3. Scheme of experiment Exp. No. Speed (RPM) Feed
(mm/min)
Allowance (mm)
1 600 80 0.2
2 600 120 0.3
3 600 160 0.4
4 800 80 0.3
5 800 120 0.4
6 800 160 0.2
7 1000 80 0.4
8 1000 120 0.2
9 1000 160 0.3
The levels are taken with the help of range of the variable parameters are chosen with the help of GUHRING international manual for advanced machine tools. Each test is repeated three times in order to achieve validity and accuracy. The experiments were performed with TiN coated carbide reamer tool perpendicular to the work surface. The L9 array with all values selected for the experimental run is shown in table 3.
Analysis of Variance
The analysis of variance (ANOVA) is a common statistical technique to significant factor that we are considered for experimentation. It calculates parameters such as sum of squares, degree of freedom, variance and percentage of contribution of each factor/interaction to the total variation. The sum of squares is the measure of the deviation of experimental data from mean value of the data. ANOVA is the procedure with considerable statistical formula which is very complicated. The Fisher’s ratio is also called as F value. F value is the ratio of mean square to the error. The principle of the F test is that, the larger value for a particular factor (parameter), the greater the effect on the response due to the change in that factor. Compare the f value obtained with that from the f distribution table. Those values obtained experimentally more than that from the f distribution table will represent the corresponding factor which is most significant. From the ANOVA table the percentage contribution of each factor can also be found.
Signal to Noise Ratio
Adoption of Taguchi method of design of experiment is to reduce the number of experiments, yet cover the entire parameter space with the help of a special design of orthogonal array. The results of such experiments are then transformed into a signal to noise (S/N) ratio to measure the deviation of the performance characteristics from the desired values. In the analysis of S/N ratio, the categories of performance characteristics selected are as follows:
Smaller the best: S/N = -10log
∑ ( )
Where n is the number of measurements in a trial and yi is the measured value in a trial.
Nominal the best: S/N = 10log
Where,
=
.………… ,=
∑ ( )Where n is the number of measurements in a trial and yi is the measured value in a trial, µ is the mean and σ is the variance.
Larger the best: S/N = -10log
∑
Where n is the number of measurements in a trial and yi is the measured value in a trial.
Grey Relational Analysis
Grey relational analysis is used for solving interrelationships among the multiple responses. In this approach a grey relational grade is obtained for analysing the relational degree of the multiple responses.
Optimization Steps in Grey Relational Analysis
Step 1: Transform the original response data into S/N ratio using appropriate formulae depending on the type of quality characteristic.
Step 2: Normalize Yij as Xij (0≤ Xij≤1) by the following formula to avoid the effect of using different units and to reduce variability. Normalization is a transformation performed on a single input to distribute the data evenly and scale it into acceptable range for further analysis.
Xij = Normalized value for ith experiment/trial for jth dependant variable/response
Larger the better :
Smaller the better :
Where yij is the ith performance characteristic in the jth experiment .maxi y ij and mini yij are the maximum and minimum values of ith performance characteristic for alternate j, respectively.
Step 3: Compute grey relational coefficient (GC) for the normalized S/N ratio values.
GCij = (Δmin+ λΔmax)/(Δij+ λΔmax)
Where, GCij = grey relational coefficient for the ith experiment/trial and jth dependant variable/response
Δ = absolute difference between Yoj and Yij which is a deviation from target value and can be treated as quality loss.
Yoj = optimum performance value or the ideal normalized value of the jth response, Yij = the ith normalized value of the jth response/dependant variable
Δmin = minimum value of Δ, Δmax = maximum value of Δ
λ is the distinguishing coefficient which is defined in the range 0≤ λ ≤1 ( the value may be adjusted on the practical needs of the system)
Step 4: Compute the grey relational grade (Gi); Gi = (1/m)Σ GCij, Where m is the number of responses
Step 5: The grey grade is equivalent to MRPI and is treated as single response problem and MRPI data can be analysed to determine the optimal level for the factors. Also response graph method or ANOVA can be used to select optimal levels for the factors based on maximum average Gi value.
Experimental Procedure
3 blocks of SS 440C with dimensions 55mm length, 45mm width and 30mm thickness were selected. In each block 9 holes with 4.5mm holes with diameter and 30 mm height were done with the help of special purpose vertical milling machine using TiN coated carbide reamer. Identification number starting from A1 to A9, B1 to B9 and C1 to C9 were engraved on the side of the work piece for tracking.
Output responses
The main output responses considered for optimization of reaming process on SS 440C using TiN coated carbide reamer are surface finish and cylindricity.
Surface Finish
By definition, surface finish is the allowable deviation from a perfectly flat surface that is made by some manufacturing process. All machining processes will produces some roughness on the surface. This roughness can be caused by a cutting tool, cutting rate and environmental conditions and the type of material for working with. Surface finish is generally broken up into three components such as roughness, waviness, and form. Roughness is generally the machined marks made on a surface by the cutting tool. Waviness is the result of the vibration of the tool. Form surface irregularities caused by worn off machine bad, table etc. All three surface finish components exist simultaneously. They simply overlap one another. We often look at each (roughness, waviness, and form) separately, so we make the assumption that roughness has a shorter wavelength than waviness, which in turn has a shorter wavelength than form.
The average roughness is the area between the roughness profile and its mean line, or the integral of the absolute value of the roughness profile height over the evaluation length. Surface roughness average (Ra), also known as arithmetic average (AA) is rated as the arithmetic average deviation of the surface valleys an peaks expressed in micro inches or micro meters. ISO standards use the term CLA (Center Line Average). Both are interpreted identical. The equation for average surface roughness can be calculated using following equation.
Ra = (h1+h2+---+hn)/n
Where, n is the number of peaks and valleys and hi, the measured value of each peaks and valleys
Figure 7: Average surface roughness profile
The measurements are usually made along a line, running at right angle to the general direction of tool marks on the surface.
Actual profile(Af) is the profile of the actual surface obtained by finishing operation.
Reference profile (Rf) is the profile to which the irregularities of the surface is referred to. It passes through the highest point of the actual profile.
Datum profile (Df) is the profile, parallel to the reference profile .it passes through the lowest point B of the actual profile
Actual profile (Af) is the profile of the actual surface obtained by finishing operation.
Reference profile (Rf) is the profile to which the irregularities of the surface is referred to. It passes through the highest point of the actual profile.
Datum profile (Df) is the profile, parallel to the reference profile. It passes through the lowest point B of the actual profile.
Cylindricity
Cylindricity is a tolerance of form that checks cylindrical features only. All circular and longitudinal elements on the surface of the cylinder must lie within two concentric cylinders separated by the value of the cylindricity tolerance which is measured on radius. Cylindricity requires that the entire face of the cylinder be contained between two concentric cylinders. Cylindricity is represented by ‘ ’.
Figure 8: Cylindricity representation
The cylindricity of the part in this instance affects directly on the efficiency, safety, and life of an assembly.
III. RESULTSANDDISCUSSION
Results were obtained for surface finish and cylindricity for reaming of SS 440C experimental blocks. MINITAB 16 software is used to perform Taguchi design of experiment and ANOVA. Grey Relational Analysis was performed to found the optimum combination of input parameters.
Analysis of Experimental Data
Statistical analysis of variance (ANOVA) is done to investigate which design parameter significantly affects the surface finish and cylindricity. Based on the ANOVA, the relative importance of the reaming parameters with respect to surface finish and cylindricity were investigated to determine the percentage contribution of each factors. All analysis is carried out for a significance level of α=0.05, i.e., for confidence level of 95%. ANOVA table also has probability level that is the realized significance level, associated with the F-tests for each source of variation. The sources with a probability level less than 0.05 are considered to have a statistically significant contribution to the performance measures.
Also the percentage of contribution of each source to the total variation indicates the degree of influence on the result by each source.
Taguchi’s method of analyzing means of the S/N ratio using conceptual approach involves graphical method for studying the effects and visually identifying the factors that appear to be significant. The rank indicates the dominant machining parameter.
Experimental Combinations
Three blocks of SS 440C where chosen for conducting experiments. In each block, 9 holes were reamed based on experimental combinations obtained. All the trials where repeated thrice in order to validate the experimental results.
Figure 9 shows the experimental blocks chosen for this work.
Figure 9: Experimental Blocks Table 4: Reaming experimental combinations Trial
No.
Bore No.
Speed (RPM)
Feed (mm/min)
Allowance (mm)
1 A1, B1, C1 600 80 0.2
2 A2, B2, C2 600 120 0.3
3 A3, B3, C3 600 160 0.4
4 A4, B4, C4 800 80 0.3
5 A5, B5, C5 800 120 0.4
6 A6, B6, C6 800 160 0.2
7 A7, B7, C7 1000 80 0.4
8 A8, B8, C8 1000 120 0.2
9 A9, B9, C9 1000 160 0.3
Experimental Results for Different Trials
The experimental results obtained by reaming experiment conducted on three blocks of SS 440C were tabulated below on table 5. The responses investigate are surface finish and cylindricity. The experiment was conducted three times, in order to achieve repeatability. The ranges for different parameters are set with reference to Guhrings international manual for special tools.
Table 5: Experimental results for different trials
Trial No.
Surface Roughness, SR
(µm) Cylinricity, (µm)
SR1 SR2 SR3 1 2 3
1 0.59 0.61 0.59 5.3 5.32 5.32
2 0.51 0.53 0.52 6.15 6.21 6.19
3 0.54 0.54 0.55 4.34 4.34 4.35
4 0.57 0.59 0.58 7.3 7.36 7.34
5 0.53 0.53 0.52 4.08 4.08 4.09
6 0.51 0.51 0.5 2.55 2.55 2.5
7 0.69 0.71 0.7 5.06 5.1 5.09
8 0.54 0.54 0.53 4.13 4.13 4.12
9 0.51 0.53 0.51 5.24 5.32 5.31
Grey Relational Analysis
Table 6: Experimental data for grey relational analysis Trial
No.
Surface Roughness, SR
(µm) Cylinricity, (µm)
SR1 SR2 SR3 1 2 3
1 0.59 0.61 0.59 5.3 5.32 5.32
2 0.51 0.53 0.52 6.15 6.21 6.19
3 0.54 0.54 0.55 4.34 4.34 4.35
4 0.57 0.59 0.58 7.3 7.36 7.34
5 0.53 0.53 0.52 4.08 4.08 4.09
6 0.51 0.51 0.5 2.55 2.55 2.5
7 0.69 0.71 0.7 5.06 5.1 5.09
8 0.54 0.54 0.53 4.13 4.13 4.12
9 0.51 0.53 0.51 5.24 5.32 5.31
Grey relational analysis is used for solving interrelationships among the multiple responses. In this approach a grey relational grade is obtained for analysing the relational degree of the multiple responses. The first step is to transform the original response data into S/N ratio using appropriate formulae depending on the type of quality characteristic given previously. The next step is to normalize Yij as Xij (0≤ Xij≤1) by the appropriate formula to avoid the effect of using different units and to reduce variability. Normalization is a transformation performed on a single input to distribute the data evenly and scale it into acceptable range for further analysis.
Table 7: S/N ratios and Normalized S/N ratios for grey relational analysis Trial
No.
S/N ratios Normalized S/N ratios
SR SR
1 4.48428026 -14.5073549 0.506066 0.696841
2 5.67886251 -15.8245239 0.080604 0.839517
3 5.2983461 -12.7564683 0.216129 0.507184
4 4.73057954 -17.3060788 0.418344 1
5 5.56893543 -12.2203025 0.119756 0.449106
6 5.9051774 -8.07422268 0 0
7 3.09744836 -14.1230203 1 0.655209
8 5.4055725 -12.3119935 0.177939 0.459038
9 5.73434523 -14.46931 0.060844 0.69272
Table 8: Grey grade calculation for grey relational analysis
Quality Loss Grey relational Coefficient Grey grade value
ΔSR Δ/O/ GCSR GC Gi
0.493934 0.303159 0.669374 0.767366 0.71837
0.919396 0.160483 0.520997 0.86171 0.691354
0.783871 0.492816 0.560579 0.669875 0.615227
0.581656 0 0.632249 1 0.816124
0.880244 0.550894 0.531846 0.644789 0.588318
1 1 0.5 0.5 0.5
0 0.344791 1 0.74361 0.871805
0.822061 0.540962 0.548829 0.648945 0.598887
0.939156 0.30728 0.515688 0.764947 0.640317
Compute grey relational coefficient (GC) for the normalized S/N ratio values. After normalized S/N ratio found quality loss. Compute grey relational coefficient and thereafter grey grade as seen from table 14 using appropriate formula given previous sections.The grey grade is equivalent to Multi Response Performance Index (MRPI) and is treated as single response problem and MRPI data is analyzed to determine the optimal level for the factors. Mean MRPI values are tabulated on table 9.
Table 9: Mean MRPI values
Factors Levels
1 2 3
Speed 0.674983 0.634814 0.70367
Feed 0.8021 0.626186 0.585181
Allowance 0.605752 0.715932 0.691783
Figure 10: Factor effects on mean MRPI
From the table 9, it is found that the optimal levels are A3, B1 and C1
Graphical representation showing the factor effects on mean MRPI are shown in figure 10.
The optimal combination found from effect on mean MRPI are:
Speed = 1000 RPM Feed = 80 mm/min Allowance = 0.3 mm ANOVA for Grey Grade
This is done to find the factors which are significant and to find the percentage contribution of each factors in attaining the responses surface finish and cylindricity together. The main effects plot for S/N ratio is shown in figure 11.
Table 11: S/N ratio for grey grade value
Trial No. Grey Grade Gi S/N ratio for Gi
1 0.7184 -2.873
2 0.6914 -3.2059
3 0.6152 -4.2193
4 0.8161 -1.7649
5 0.5883 -4.6078
6 0.5 -6.0206
7 0.8718 -1.1916
8 0.5989 -4.4531
9 0.6403 -3.8721
0 0.2 0.4 0.6 0.8 1
1 2 3
Speed 0.674983415 0.634814005 0.703669891
Feed 0.802099773 0.626186156 0.585181382
Allowance 0.605752312 0.715931903 0.691783096
Mean MRPI
Factor Effects on Mean MRPI
1 0 0 0 8 0 0
6 0 0 -2 -3 -4
-5
1 6 0 1 2 0
8 0
0 .4 0 .3
0 .2 -2 -3 -4 -5
Spe e d
Mean of SN ratios
Fe e d
A llo w ance
Main Effects P lot for S N r atios Data M eans
S ignal-to-noise: Larger is be tter
Figure 11: Main effects plot for S/N ratio for grey grade Table 12: Response Table for grey grade S/N ratios (Larger is better)
Level Speed Feed Allowance
1 -3.433 -1.943 -4.449
2 -4.131 -4.089 -2.948
3 -3.172 -4.704 -3.34
Delta 0.959 2.761 1.501
Rank 3 1 2
Table 13: Analysis of Variance for grey grade
Source DOF Sum of
squares
Mean of
squares F P contribution
(%)
Speed 2 1.4748 0.7374 7.75 0.114 8.24
Feed 2 12.605 6.3024 66.25 0.015 70.38
Allowance 2 3.638 1.819 19.12 0.05 20.32
Error 2 0.1903 0.0951 1.06
Total 8 17.908 100
Based on the ANOVA results in Table 13 the percentage contribution of various factors to cylindricity is identifiable.
Here, feed and allowance are the most influencing factor followed by speed. The percentage contribution of allowance and feed towards grey grade are 70.38% and 20.32% respectively.
The optimal combinations found from main effects plot for S/N ratio are:
Speed = 1000 RPM Feed = 80 mm/min Allowance = 0.3 mm Regression equation
The regression equation is
Grey Grade = 0.810 + 0.000072xSpeed – 0.00271xFeed + 0.430xAllowance Confirmation Test
The confirmation test is the final step in verifying the results obtained from Taguchi’s design approach. The optimal conditions are set for the significant factors and experiments are run under specified reaming conditions. The results from the confirmation experiment are compared with the predicted average based on the parameters and levels tested. The confirmation experiment is a crucial step and is highly recommended by Taguchi to verify the experimental results.
The purpose of the confirmation experiment is to validate the conclusions drawn from the experiment. Confirmation experiment is conducted using the optimal levels of the significant factors. For the insignificant factors, although any level can be selected, but I selected according to the optimum combination obtained using grey relational analysis. The confirmation test block is shown in figure 12.
Figure 12: Confirmation test block
The confirmation test is conducted on SS 440C test block with TiN coated carbide reamer with the optimum level of combinations found using grey relational analysis. The confirmation test results were tabulated in table 14. In the confirmation test, the responses for which data collected are for surface roughness and cylindricity. Comparison with the existing process was also done.
Table 14: Confirmation test results
Bore No. Surface Roughness, SR (µm) Cylindricity, (µm)
1 0.26 2.47
2 0.32 2.55
3 0.27 3.03
4 0.28 2.85
5 0.23 2.48
Validating Confirmation Test
It is important to validate the experimentally confirmed results with the statistically obtained results. For that the regression equation obtained for grey grade is used.
Regression equation obtained for grey grade is;
Grey Grade = 0.810 + 0.000072xSpeed – 0.00271xFeed + 0.430xAllowance
It is used optimum combination that is speed (1000 RPM), feed (80 mm/min) and allowance (0.3 mm) for the confirmation experiment. That is; Grey grade = 0.810 + (0.000072 x 1000) – (0.00271 x 80) + (0.430 x 0.3) = 0.7942
Table 15: S/N ratio and Normalized S/N ratio for confirmation test Bore
No.
Surface Roughnes, SR
(µm)
Cylindricity, (µm)
S/N ratio for SR
S/N ratio for
Normal- ized SR
Normal- ized
1 0.26 2.42 11.7005 -7.6763 1 0
2 0.24 2.44 12.3958 -7.7478 0.3471 1
3 0.26 2.43 11.7005 -7.7121 1 0.501
4 0.25 2.44 12.0412 -7.7478 0.6801 1
5 0.23 2.43 12.7654 -7.7121 0 0.501
Grey grade calculation for confirmation test results: Grey grade calculation for the confirmation test was done for comparing it with the value obtained by substituting the optimum combinations of parameters obtained. For that, the same procedure that explained in chapter 3 need to be done. First find out the S/N ratio, normalized S/N ratio, grey grade coefficient and grey grade.
Table 16: Grey grade calculation for confirmation test results
Bore No. Quality Loss Grey relational
Coefficient Grey grade value
ΔSR Δ GCSR GC Gi
1 0 1 1 0.5 0.75
2 0.6529 0 0.605 1 0.8025
3 0 0.499 1 0.6671 0.8336
4 0.3199 0 0.7576 1 0.8788
5 1 0.499 0.5 0.6671 0.5836
The grey grade obtained for different bores done for confirmation test are tabulated on table 16. This value can be compared with the value obtained when optimum values of parameters when applied on regression equation. The graphical representation for this comparison is shown in figure 13.
Figure 13: Grey grade comparison
From this comparison it is seen that the grey grade obtained by the experimental results and the value obtained by regression equation. The average value for the confirmation tests is 0.7697, whereas grey grade obtained by regression equation is 0.7942.
Surface topography of Bore Wire EDM & Reaming
Taylor Hobson digital surface profilometer images are taken to analyze the roughness and other surface characteristics to compare different surface topography obtained by the existing wire edm process and reaming process.
Scanned surface gave a rough idea about the surface characteristics. Figure 14a shows the scanned image of Wire EDMed surface and figure 14b shows the scanned image of reamed surface where TiN coated carbide reamer was used as the reaming tool. From these images, the surface integrity of WEDM surface can be easily identified.
White layers occurred on the WEDMed surface can be seen on the scanned image.
Figure 14: Scanned surface a) WEDMed surface, b) Reamed surface
Figure 15: 2D view of a) WEDMed surface b) Reamed surface,
0 0.5 1
0 5 10
Grey Grade
Bore Number
Grey Grade Comparison
Series1 Series2
3D view of c) WEDM surface, d) Reamed surface
Figure 16: Flattened 3D view of a) WEDMed surface, b) Reamed surface
The parallel swatches obtained by the reaming process can also be seen from the scanned images. These images are taken with the help of Taylor Hobson digital surface profilometer for the further investigation on the surface characteristics of the surfaces. Figure 15 shows the 2D and 3D representation of bore manufactured by WEDM and reaming process. Flattened 3D images of bore surface to get the overall idea about the surface texture are shown in the figure 16.
Comparison of Cylindricity
Figure 17: a) Cylindricity of WEDMed surface with 1µm/div, b) Cylindricity of reamed surface with 1µm/div, c) Cylindricity of WEDMed surface with 5µm/div, d) Cylindricity of reamed surface with 5µm/div
Cylindricity of the existing WEDM process and reaming process where compared with the help of TAYLOR HOBSON TALYROND 365. The images obtained are shown in figure 17.From the figure 17, it is easy to compare the cylindricity obtained by the existing WEDM process and reaming process. While considering the manufacturing by WEDM process, honing and lapping is essential to attain the required cylindricity, but by using reaming process, at least one process that is honing can be avoided because of lower cylindricity value obtained.
Data Comparison for Existing and Proposed Process
While we comparing the existing and proposed operation cycles in the manufacturing the spool bore of valve body, the main difference is by the replacing currently using WEDM process with reaming process.
Table 17: Responses for bore manufacturing using existing operation cycle
Compt. ID No.
Cylindricity Value Surface Roughness Value
After Wire EDM (µm)
After Honing (µm)
After Lapping (µm)
After Wire EDM (µm)
After Honing (µm)
After Lapping (µm)
A 1 7.19 3.52 2.08 0.84 0.35 0.2
A 2 8.17 3.49 2.23 0.86 0.42 0.22
A 3 9.27 3.46 2.14 0.68 0.34 0.18
A 4 7.31 3.28 2.01 0.76 0.37 0.2
A 5 8.81 3.08 1.72 0.68 0.3 0.08
The responses taken are obtained by each operation cycle are cylindricity and surface roughness. The response values obtained while the spool bore manufactured by the existing operation cycle are as tabulated on table 4.19
The tabulated values are obtained by analyzing a batch of EHSV. After WEDM, after honing and after lapping processes are tabulated on the table 17
For comparing these values with proposed operation cycle, the response values obtained while the spool bore manufactured by the proposed operation cycle are as tabulated on table 4.20.
The tabulated values are obtained by analyzing the results obtained from the confirmation test. After reaming and after lapping processes are tabulated on the table 18.
Table 18: Responses for bore manufacturing using proposed operation cycle
Bore No.
Cylindricity Value Roughness Value After
Reaming (µm)
After Lapping (µm)
After Reaming (µm)
After Lapping (µm)
1 2.47 1.81 0.26 0.12
2 2.55 1.68 0.32 0.22
3 3.03 2.01 0.27 0.11
4 2.85 1.89 0.28 0.13
5 2.48 2.03 0.23 0.08
By comparing the values obtained by existing operation cycle and proposed operation cycle, it is seen that by implementing the reaming process to the existing operation cycle by replacing WEDM. It is found that by adopting reaming process instead of WEDM will help to avoid at least one process that is honing.
IV. CONCLUSION
From the experiments, it was found that the surface finish was significantly influenced by feed; however speed and allowance has very less effect while machining SS 440C using TiN coated carbie reamer. Using Grey Relational Analysis, the optimal combination was obtained at speed (1000 RPM), feed (80 mm/min), and allowance (0.3mm). This optimum combination is used for confirmation experiment. This confirmation experiment was validated by analyzing the regression equations developed with the confirmation test results.
ACKNOWLEDGMENT
First and the foremost, I thank God, the Almighty who gave me the inner strength, resource and ability to complete the work successfully without which all the efforts would have been in vain. I express my heartfelt thanks to the Head and Professor, Department of Mechanical Engineering and my internal guide, Prof. Dr. Josephkunju Paul C from Mar Athanasius College of Engineering, Kothamangalam, Kerala, Inia for his support, valuable advice and helpful feedback for the successful completion of my work. I express my sincere gratitude and I am highly grateful to my external guide Dr. S. Karunanidhi, Scientist G and Deputy Director, Control Systems Laboratory, Research Centre Imarat, Defense Research and Development Organization (RCI DRDO), Hyderabad for giving me this golden opportunity to be a part of the organization, encouraging supervision, valuable discussions, never failing kindness and inspiring guidance throughout the work.
REFERENCES
[1] F. Klocke, M. Zeis, A. Klink, D. Veselovac, 2013, Technological and economical comparison of roughing strategies via milling, sinking-EDM, wire-EDM and ECM for titanium and nickel based blisks, CIRP Journal of Manufacturing Science and Technology, 198-203.
[2] D. Welling, 2014, Results of surface integrity and fatigue study of Wire-EDM compared to broanching and grinding for demanding jet engine components made of Inconel 718, Procedia CIRP, 339-344.
[3] L. Li, Y.B. Guo, X.T. Wei, W. Li, 2013, Surface integrity characteristics in wire-EDM of inconel 718 at different discharge energy, Procedia CIRP, 220-225.
[4] F. Klocke, D. Welling, J. Dieckmann, 2011, Comparison of grinding and Wire EDM concerning fatigue strength an surface integrity of machined Ti6Al4V components, Procedia Engineering, 184-189.
[5] S. Karunanidhi, M. Singaperumal, 2010, Design, analysis and simulation of magnetostrictive actuator and its application to high dynamic servo valve, Sensors and Actuators, 185-197.
[6] Fuzhu Han, Jie Zhang, Isago Soichiro, 2007, Corner error simulation of rough cutting in wire EDM, Precision Engineering, 331-336.
[7] D.K. Aspinwall, S.L. Soo, A.E. Berrisford, G. Walder, 2008, Workpiece Surface roughness and integrity after WEDM of Ti-6Al-4V and Inconel 718 using minimum damage generator technology, CIRP Annals – Manufacturing Technology, 187-190.
[8] Haruki Obara, Harutoshi Satou, Masatoshi Hatano, 2004, Fundamental study on corrosion of cemented carbide during wire EDM, Journals of Material Processing Technology, 370-375.
[9] K.H. Ho, S.T. Newman, S. Rahimifard, R.D. Allen, 2004, State of the art in wire electric discharge machining (WEDM), International Journal of Machine Tools & Manufacture 1247-1259
[10] B. Bojorquez, R.T. Marloth, O.S. Es-Said, 2002, Formation of a crater in the workpiece on an electric discharge machine, Engineering failure Analysis, 93-97.
[11] P. Muller, G. Genta, G. Barbato, L. De Chiffre, R. Levi, 2012, Reaming process improvement and control: An application of statistical engineering, CIRP Journal of Manufacturing Science and Technology, 196-201
[12] J. Pradeep Kumar, P. Packiaraj, 2012, Effect of drilling parameters on the surface roughness, tool wear, material removal rate and hole diameter error in drilling of OHN, International Journal of Avanced Engineering Research and Studies, 150-154
[13] Leonardo De Chiffre, Guide Tosello, Miroslav Piska,Pavel Muller, 2009, Investigation on capability of the reaming process using minimal quantity lubrication, CIRP journal of Manufacturing Science and Technology, 47-54.
[14] GUHRING International Manual on advanced machine tools, 9th edition, 2006.
[15] K. Krishnaiah, P. Shahabudeen, 2012, Applied design of experiments and Taguchi methods, PHI learning Private Limited,
[16] H. Dagnall M.A., 1996, Let’s Talk Roundness, Taylor Hobson Limited.
[17] Papers on Precision engineering, SERC School on ‘Precision Engineering’, 2001, Indian Institute of Technology Madras, Chennai, December 4-14
[18] R. Paneerselvam, 2013, Research Methodology, PHI Learning Private Limited.
[19] R. Paneerselvam, 2012, Production and Operations Management, PHI Learning Private Limited.
